Skip to main content

Introduction to Ravioli

Welcome to the Ravioli documentation site!

Ravioli is a modern, AI-Native, privacy-respecting personal data warehouse and "Vibe-Analytics" platform built for agile teams and individuals. Part of the AI Passione ecosystem, Ravioli combines professional-grade data engineering with an interactive notebook interface, running fully local on DuckDB with seamless integrations to MotherDuck, Ollama, and Notion.


Core Pillars & Documentation Guide

Explore the architectural components and features of the Ravioli platform:

📊 Insights

Learn about the formal review lifecycle of analytical facts. Trace observed data points to final team conclusions using the:

  • Feed & Summary: Chronological aggregation feed of verified platform findings.
  • Lineage Map: Relational directed graph showing parent-child insight links.

📈 Analyses

Explore the interactive playgrounds where data exploration occurs:

  • Quick Insights: Instant statistical profiling generated automatically on data upload.
  • Custom Notebooks: Cell-based execution journals saving queries and rendering outputs via SSE streaming.
  • Deep Dives (Upcoming): AI-driven autonomous query corrections and dynamic visualization chart layouts.

📓 Knowledge Base

Manage company-specific context definitions to ground AI analysts and prevent hallucinations. Learn how pages are stored as Notion-style block lists and sync bi-directionally.

💾 Data

Understand how raw observations are brought into the local warehouse and structured:

  • Data Ingestion: Pipelines for flat files (CSV, Parquet, JSON, GPX), API endpoints (WFS geospatial layers), and upcoming DLT connectors.
  • Data Transformation (ECL): The AI-powered Extract, Contextualize, and Load pipeline featuring PII masking and semantic schema inferences.

🛡️ Governance

Maintain absolute integrity over analytical outputs. Read about data ownership, stewardship workflows, and access control models:

  • Insights Review: Auditing steps by Admins and Stewards.
  • Users & Groups: User profiles, team group structures, and lineage attribution tags.

🔌 Settings & Integrations

Configure user profiles, manage personal prompts, and link external systems:

  • Basic Settings: Contact metadata and customized persona rules.
  • Integrations: Local/cloud LLMs (Ollama, Gemini), Knowledge tools (Notion, Confluence), and Cloud Warehouses (MotherDuck, BigQuery).
  • Notion Sync Guide: Deep dive into Notion's API block structures and sync intervals.